Hi all,

We need your opinion

What mechanism do you prefer to use the YAML Stacks Parser ? ( https://github.com/jboss-jdf/jdf-stack)
* Copy the existing Parser code to my project
*
Add a maven dependency that provides the Parser to my pom.xml

It takes just some seconds to vote on http://www.surveymonkey.com/s/X78GDG3 (Due date: 08/24/2012)

Thanks

-------- Mensagem original --------
Assunto: Re: [jdf-dev] Stacks Client
Data: Wed, 22 Aug 2012 16:53:15 -0300
De: Rafael Benevides <benevides@redhat.com>
Empresa: Red Hat
Para: jdf-dev@lists.jboss.org, Fred Bricon <fbricon@redhat.com>, Pete Muir <pmuir@redhat.com>, James Perkins <jperkins@redhat.com>


Today Fred Bricon asked about using the YAML Stacks file Parser that is available on Stacks Repo.

It was suggested to have a java project so it will be possible to understand the Parser dependencies and how it should be used.
For that, it was created the following issue: https://issues.jboss.org/browse/JDF-98

However we already have this implementation that we could promote as an official Stacks client: https://github.com/jamezp/jdf-stack-client and merge it on Stacks Repo

I'm resurrecting this Thread so we can discuss again and vote.

To help us, I've created the following Poll http://www.surveymonkey.com/s/X78GDG3 (no signup needed)

Please, give your opinion on this subject until 08/24/2012.

Thanks


Em 13-08-2012 14:56, James Perkins escreveu:
Just one thing to add about a dependency too. Since the client wouldn't 
really be a huge project, using the shade plugin would be easy enough 
for anyone that doesn't want a dependency.

On 08/13/2012 08:31 AM, James Perkins wrote:
Most of the code was from that Parser. Rafael really deserves the credit
for the code :-) I just copied it and I just refactored a bit was all. I
only renamed the Runtime class to ServerRuntime because IDEA kept
annoying me thinking I was trying to use java.lang.Runtime.

I normally don't like adding dependency, but in this case one needs to
be added for a YAML parser anyway. Adding one more to me isn't really a
big deal if it's more convenient.

On 08/13/2012 03:35 AM, Pete Muir wrote:
Hey James,

My initial proposal was to create a "reference client", that would be really simple. Rafael implemented it in ~550 LOC - https://github.com/jboss-jdf/jdf-stack/blob/master/Parser.java as I know that adding dependencies can get really complex (e.g. consider someone using JBDS and Forge together, and both using different versions of the client. Then this reference client can be copied in, and used.

However, obviously we really want to do what makes most sense for the people who will use the client:

* JBoss AS Maven plugin
* JBoss AS Forge plugin
* JBoss AS Eclipse plugin
* Arquillian (not discussed this one so far, but probably useful ;-)

So, I guess we should call a vote :-)

Who would prefer a dependency over a copy and paste? Please respond!

On 11 Aug 2012, at 01:00, James Perkins wrote:

I've been working on getting stacks integrated with the maven plugin for the run goal. While doing this I got to wondering if we should have a stacks client. I believe Pete may have mentioned something about not wanting one, but it just seems we'll end up with a bunch of duplicate code.

I've created a quick prototype https://github.com/jamezp/jdf-stack-client. It would need some work and could probably be made a little more generic for integrating with tooling, but it's just a PoC so far.

Any thoughts or opinions about having a client available?
-- 
James R. Perkins
JBoss by Red Hat

_______________________________________________
jdf-dev mailing list
jdf-dev@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/jdf-dev